UK government advisers warn Britain risks becoming AI customer rather than creator

The UK Council for Science and Technology has warned Minister for AI Kanishka Narayan that Britain risks becoming a high-value AI customer rather than a technology leader without deliberate action. To protect strategic autonomy against dominant U.S. technology firms, the advisory body recommends creating a national AI Sovereignty Plan, appointing a single government lead, investing in low-power computing and dataset infrastructure, and launching public sector pilots across health, education, and national security.

Context of the Council's letter

• ▪Council for Science and Technology co-chairs Dame Angela McLean and Lord Browne of Madingley addressed their recommendations for a UK AI Sovereignty Plan in a letter copied to Prime Minister Keir Starmer and Minister for AI Kanishka Narayan
• ▪The UK Council for Science and Technology issued a letter warning that Britain risks becoming a high-value AI customer rather than a creator unless it takes deliberate action to build domestic capabilities

Governance and strategic planning recommendations

• ▪The Council for Science and Technology recommended that the UK government develop an official AI Sovereignty Plan mapping computing power, data, skills, models, governance, and public adoption
• ▪The Council for Science and Technology recommended appointing a single accountable government lead, such as the head of the Cabinet Office AI Taskforce, to coordinate the UK's AI sovereignty policy

Public interest AI pilots

• ▪The Council for Science and Technology specified six criteria for every pilot under its proposed AI Sovereignty Plan, requiring assessment of sovereign data held, problem definition, model capability, system architecture, value capture, and success metrics
• ▪The UK Council for Science and Technology recommended running rapid public interest AI pilots in national security, health, and education

Investment in domestic computing capabilities

• ▪The Council for Science and Technology advised investing directly in UK computing infrastructure, skills, and data management to secure national autonomy
• ▪The Council for Science and Technology recommended supporting fundamental computing research, specifically focusing on low-power computing approaches

Defining and establishing AI sovereignty

• ▪The Council for Science and Technology defines AI sovereignty as maintaining agency across the value chain to capture economic value, deliver public outcomes, and maintain resilience under duress
• ▪The Council for Science and Technology stated that the foundations of artificial intelligence sovereignty encompass people, skills, organizations, data, trust, and adoption alongside models

Developing AI skills and tracking adoption

• ▪The Council for Science and Technology recommended long-term investment in UK AI skills and a secondment route to complement the existing TechFirst program
• ▪The Council for Science and Technology recommended building a stronger evidence base regarding artificial intelligence adoption and productivity in the United Kingdom

Developing domestic AI models and autonomy

• ▪The Council for Science and Technology emphasized that decisions made now on AI sovereignty will shape the United Kingdom's long-term competitiveness, resilience, public trust, and strategic autonomy
• ▪The Council for Science and Technology proposed exploring a publicly governed British AI model made available to researchers, startups, and public services

Building domestic AI development and research

• ▪The Council for Science and Technology warned that skills, commercial value, and technological capability will accumulate primarily where AI systems are directly developed and used
• ▪The Council for Science and Technology noted that the United Kingdom is well positioned to benefit from the UK's strong research base and trusted institutions

Managing and accessing national datasets

• ▪The Council for Science and Technology proposed identifying key British datasets and testing access arrangements in at least two high-value sectors
• ▪The Council for Science and Technology emphasized that AI sovereignty cannot be achieved through isolated technology investments and requires managing datasets across high-value areas like health and creative industries

Balancing sovereign development and global access

• ▪The Council for Science and Technology stated that the United Kingdom must determine which technologies to develop domestically and which to access securely from other nations
• ▪The Council for Science and Technology advised the UK government to maintain access to leading global AI models while selectively developing sovereign capability to avoid universal external dependence